960 Revised Lille Score for Predicting 90 Days Mortality in Severe Alcoholic Hepatitis

Abstract

INTRODUCTION: Iron is an essential micronutrient which circulates bound to transferrin for tissue distribution. Liver is the main site for transferrin synthesis and its dysfunction can alter iron homeostasis resulting in oxidative stress. Recent studies showed that transferrin saturation is an independent predictor of mortality in patients with decompensated cirrhosis. We aim to evaluate the prognostic efficacy of transferrin saturation and revised Lille score to predict 90-days mortality in patients with severe alcoholic hepatitis. METHODS: We retrospectively reviewed the electronic medical records of patients diagnosed with severe alcoholic hepatitis (SAH) who received prednisolone therapy. Demographic and biochemical data at diagnosis and day 4 was collected. Univariate logistic regression for all variables included in Lille score along with ferritin and transferrin percent was performed. A revised Lille score was created based on results of univariate analysis. The efficacy of revised Lille score was then tested against Lille score on day 4 with receiver operating characteristics (ROC) analysis and Kaplan-Meier survival analysis by dichotomization per optimal cut-offs. RESULTS: In our cohort of 80 patients with SAH, 12(13.3%) patients died within 90 days. Apart from all parameters included in the Lille score, transferrin saturation was identified as an independent predictor of 90-days mortality on univariate analysis (Table 1). Revised Lille score was generated by the formula mentioned below. The ROC analysis for the prediction of 90-day mortality of Day 4 revised Lille score revealed a higher AUC of 0.8419 in comparison to Day 4 Lille score that had a lower AUC of 0.6985 (Table 1). With the optimal cut-off value of 5, the revised Lille score sub-classified patients into low and high-risk (HR = 21.05, P < 0.004) groups (Figure 2). Revised Lille score = 2.15 × Lille score + 0.06 × Transferrin saturation CONCLUSION: In conclusion, revised Lille score at fourth day can be utilized to predict mortality at 90 days. When using a cut-off of 5, patients can be subclassified into low and high-risk groups with a sensitivity and specificity of 97.95% and 35.48 respectively. We propose that revised Lille score should be compared with Lille score in predicting outcomes in SAH. Moreover, patients with elevated transferrin saturation have a high mortality.